334 MARINE BIOLOGY OF THE SUDANESE RED SEA. 
these fibres is very irregular, and the diameter of the holes varies from 
practically nothing to 5 mm. in diameter. Where this mesh is wide the 
sarcode coating the fibres frequently does not fill the spaces, so that large 
holes and cavities occur throughout the sponge-wall; but where the 
reticulation of the fibre is compact, the sarcode completely fills up the inter- 
spaces. Thus the sponge-wall consists of a number of tracts, of greatly 
varying size and irregular arrangement, where the sarcode around the various 
skeletal fibres has completely fused together, united to each other by similar 
tracts or by fibre-covered bars of sarcode, and between these tracts there occur 
large interspaces, transforming the wall of the sponge into a very irregular 
honeycomb. By the various spaces in the wall, the cavity of the interior of 
the sponge-cylinder is placed in communication with the exterior. 
The inner surface of the sponge-cylinder is comparatively smooth, though 
the sarcode tracts are covered with low ridges and small prominences, owing 
to the irregularity of the skeletal fibre. This fibre does not proliferate 
towards the interior of the sponge-cylinder, however, so that these irregu- 
larities of surface are confined to one plane, that of the surface itself. 
On the other hand, the outer surface of the sponge-wall is covered with a 
forest of protruding fibres, which branch out and anastomose with one another 
profusely. These fibres in their distal portions are almost bare of sarcode, 
and appear as small tree-like processes outside the general tissues of the 
sponge. 
The oscula are not easy to distinguish from the interspaces between sarcode 
tracts, but true oscula appear to be fairly common, irregularly scattered over 
the whole surface of the sponge. They are quite small, rarely exceeding 
1-0 mm. in diameter. They occur on the inside as well as the outside of the 
cylinder wall. Where pores could be distinguished, they were extremely 
numerous, scattered thickly and regularly over the dermal membrane of the 
sponge. They are quite small and appeared somewhat irregular in outline. 
They are not located in special pore-areas, and no special skeleton is present 
in connection with them. 
As stated above, the spicule-covered fibre which forms the main skeleton 
may reach a diameter of over 1:0 mm., and all sizes less than this maximum 
can be distinguished. The fibres are entirely filled up with spicules of various 
kinds, of which the most frequent by far are tylostyles, with considerable 
numbers of foreign bodies, chelee, and sigmata, and broken spicules mixed 
with them. The tylostyles and the broken megascleres, whether foreign or 
not, lie longitudinally and very regularly in the fibre, but the chelze and 
sigmata, and the non-spicular foreign bodies, which are usually quite small, 
lie entirely without orientation in the fibre. The spongin which coats the 
spicule is dull yellowish brown in colour, and even in the largest fibres does 
not greatly exceed the spicular bundle in diameter. In fact, spicules seem to 
